201224473 六、發明說明： 【發明所屬之技術領域】 本發明係關於一種電力特徵辨識裝置及其方法；更具體而言， 本發明之電力特徵辨識裝置及其方法係藉由一預設取樣次數，辨 識一電力訊號是否處於穩定狀態及辨識出該電力訊號之即時電力 特徵。 【先前技術】 基於全球環境保護與節約能源的意識高漲，能源相關議題亦逐 漸受到人們重視。能源讀表之相關應用為目前最熱門的能源應用 之一，保守估計在幾年内全球將有超過兩 億具智慧型電表（Smart meter)的換裝’以求提供使用者獲得較即時㈣電資訊。根據美 國用電資訊統計’約有39%左右的能源使用係發生在居住環境 此藉由建置先進讀表基礎建設（Advanced Metering Infrastructure ’ AMI)以提供使用者所需的用電資訊進而改變使 用者的用電彳了為將疋最重要的關鍵。*當制者充份了解自己的 用電行為，就能有效的達到減少耗電的目標。 電==態Γ係用以監視電器之使用狀態。早期， 個別安f電力Ϊ 訊’必須在每個電器 電十。隨後發展出的非侵入式的電器迴路辨_ 則只需在電力迴路上安裝一個電力計即 路辨識技冰 監視效果，更可降低電力計裝㈣數量，㈣達到㈣ 言，非侵人式的迴路辨識技術可分為電 ㈠成本。一也 段，訓練階段仙以學習—電_ ，、、階段及電器卿 将徵，而辨識階段係用 201224473 辨別所接收到電力訊號之—即時電力特徵。 於電器訓練階段，需等到電力訊號穩定後才會有良好的電力特 =入弋斤二穩定狀態係指該電力訊號變動之程度不大。習知的非 a 叫_技術’由於大多依照使用者經驗判斷各電器之 電力訊號是否 — 穩定’易造成電器訓練時間增加或是訓練失敗的可 能。此外，於 夕無、、電器辨識階段，習知的非侵入式的迴路辨識技術， 辨喊即時電力特徵，而少數可以完成辨識即時電力特徵 者，由於没右&201224473 VI. Description of the Invention: [Technical Field] The present invention relates to a power feature identification device and a method thereof. More specifically, the power feature recognition device and method of the present invention are based on a preset number of sampling times. Identify whether a power signal is in a steady state and recognize the instantaneous power characteristics of the power signal. [Prior Art] Based on the global awareness of environmental protection and energy conservation, energy-related issues are gradually gaining attention. The application of energy meter reading is one of the most popular energy applications. It is conservatively estimated that there will be more than 200 million smart meter replacements in the world in a few years to provide users with more instant (four) electricity information. . According to the US electricity consumption statistics, about 39% of energy use occurs in the living environment. This is achieved by providing the Advanced Metering Infrastructure (AMI) to provide users with the electricity information they need. The use of electricity has become the most important key to the future. * When the system fully understands its own electricity consumption behavior, it can effectively achieve the goal of reducing power consumption. The electric == state is used to monitor the state of use of the appliance. In the early days, individual power supplies must be in each electrical appliance. Subsequent development of the non-intrusive electrical circuit identification _ only need to install a power meter on the power circuit, that is, road identification technology ice monitoring effect, but also reduce the number of electricity meter installed (four), (four) reach (four), non-invasive Circuit identification technology can be divided into electrical (a) cost. In the same period, the training phase will be learned by the students, and the identification phase will use 201224473 to identify the instantaneous power characteristics of the received power signals. In the electrical training stage, it is necessary to wait until the power signal is stable before there will be a good power. The stable state of the power signal means that the power signal has not changed much. The conventional non-a _ technology is because most of the user's experience is judged whether the electrical signal of each appliance is stable - it may cause an increase in electrical training time or a training failure. In addition, in the evening, the electrical identification stage, the conventional non-intrusive loop identification technology, the instant power feature is discerned, and a few can complete the identification of the instant power feature, since there is no right &
巧民好的前處理技術，易浪費過多的計算量或是因傳 送過夕無用的封包而降低效率。 有鐘於此， 解決習知非侵 地辨識出即時 之問題。 如何建立一種電力特徵辨識裝置及其方法，以有效 入式的迴路辨識技術訓練時間不明確以及無法有效 電力特徵所造成的問題，為該領域之業者亟需解決 【發明内容】 本發明之 習知非*目的在於提供一種電力特徵辨識裝置，以有效地解決 气的迴路辨識技術訓練時間不明確以及無法有效地辨 識出即時雷夬私 乃作 力特徵所造成的問題。 為達上迷目& 的’本發明提供一種電力特徵辨識裝置，該電力特 徵辨Α裳置包 — 以持續地接收〜―器、一儲存器以及一處理器。該接收器用 收器，並用、冑力訊號。該處理器電性連接至該儲存器及該接 該處理器更^定—階段之—取樣時間間隔及—麟取樣次數。 訊，並將該對該電力訊號取樣以得該階段之一待測電力資 :’寺列電力資訊儲存於該儲存器。該處理器亦用以每隔 201224473 該取樣時間間隔對該電力訊號取樣，以各自得該階段之一參考電 力資訊，直至該等參考電力資訊之一數目等於該預設取樣次數， 並將該等參考電力資訊儲存於該儲存器。最後，該處理器用以計 算該等參考電力資訊之一統計特性，並將該待測電力資訊與該統 計特性進行比較，以得該階段之一比較結果。 為達上述目的，本發明亦提供一種用於一裝置之電力特徵辨識 方法，該裝置包含一接收器、一儲存器及一處理器，該方法包含 下列步驟：（a)令該接收器，持續地接收一電力訊號；（b)令該處理 器，設定一階段之一取樣時間間隔及一預設取樣次數；（c)令該處 理器，對該電力訊號取樣以得該階段之一待測電力資訊，並將該 待測電力資訊儲存於該儲存器；（d)令該處理器，每隔該取樣時間 間隔對該電力訊號取樣，以各自得該階段之一參考電力資訊，直 至該等參考電力資訊之一數目等於該預設取樣次數，並將該等參 考電力資訊儲存於該儲存器；（e)令該處理器，計算該等參考電力 資訊之一統計特性；（f)令該處理器，比較該待測電力資訊與該統 計特性，以得該階段之一比較結果。 為讓上述目的、技術特徵、和優點能更明顯易懂，下文係以較 佳實施例配合所附圖式進行詳細說明。 【實施方式】 以下將透過多個實施例來解釋本發明之電力特徵辨識裝置及其 方法。需說明者，由於本發明主要係涉及一用以辨識電力特徵之 裝置及其方法，因此，在以下所述的實施例及圖式中，與本發明 非直接相關之元件及步驟皆已省略而未繪示。此外，為清楚揭露 201224473 本發明之技術特徵，本案相闕圖式皆 舉係用於說㈣魏定本㈣ 為誇大时請製。此 利範圍為準。 案所請求之範圍’以申請專 本發明之第 考第1圖及第2圖。第w 线裝置】，其說明請合併參 電力迴路9之示意 =繪電力特_識裝置1應用於- 取樣示意圖。如第描繪… 11、-儲存器"及一處理器15。處特徵辨識裝置1包含-接收器 及接收51 1丨垃收 ° 處理器15電性連接至儲存器I3 ° 。器11與電力迴路9電性連結，並用以持續地 收電力迴路9上之一電力㈣？ 口 1用以持續地接 ^ 9 φ ,, ° ，電力訊號2係來自於與電力迴 路9電性連結之電器組3。 =施例主要_力特徵辨魏置ι如何於電_階段判 … 穩疋狀態，進而辨識-電器之-電力特徵。 於電㈣練階段’電力特徵辨識裝置1將分別對電器Μ、電器Μ 及電器35進行訓練’以學習電器組3之個別電器之電力特徵。以 訓練電器31為例’當電器31開啟時，接收器η自電力迴路9持 續接收電器之電力訊號2。處理器15於收到電器31之電力訊 就2钱’設I第—階段之—取樣時間間隔U—預設取樣次 數。取樣時間間隔T係用以決定每隔多少時關隔對電力訊號2 進行-次《’而預設取樣錢係心對電力訊號2每間隔該取 樣時間間m連續取樣幾次。在並非用以限定本發明之前提下，為 更明確說明本實_ ’以下將假設第—階段之預設取樣次數為4 次進行說明。 201224473 如第2圖所示，首先處理器 獲得第-階段之-待測電力資气9十^ 31之電力訊號2取樣以 ° 1，並將待測電力資訊91儲存於 儲存器13。須說明者，待測電力 ,| 冤力資sK 91於本發明之其他實施態樣 中’亦可以是每隔取樣時間間隔T對 . 次所榻取之複數個電力資訊 之訊號2取樣數 處器15將待測電力資訊91儲存 於儲存器13後，每隔取樣時間 ^ 作于間間1^ T對電器31之電力訊號2取 樣，以各自得到第一階段之— 亏電力貝汛，直至參考電力資訊 預設取樣次數。換言之，此時每取樣-次,便得- 參考電力資訊，直到得到4個參考電力f訊為止。處理器】5 一階段獲得複數個（即4個）灸 、 翏考電力資訊93，並將該等參考電 力^93儲存於儲存器13。其中，該等參考電力資訊93係包含 一參考電力資訊931、一參+f 及一參考電力資訊934。貝讯心一參考電力資訊933 將=電=器15計算該等參考電力資訊93之-統計特性，並 待測電W1與該等參考電力資訊93之統計特性進行比 較，進而得到一比較社果礙 # U舉例而舌，處理器！5可比對待測電力 疋洛入该統計特性所界定之一機率分佈範圍内，其中談 =佈：可根據該統計特性之平均值、變異數或是其他統： 订’疋’但並不褐限於此處所揭露。此處所獲得之第—产 =比Ιλ結果’將錢續程序所得之比較結果共㈣為觸電器 31是否已經完成訓練之依據。 σ 太於獲得上述比較結果後，處理器】5從第一階段之該等參考電力 資訊93中，選取參考電力資訊州，並將之設定為一第二階段之 201224473 一待測電力資訊》須說明者，於本發明之其他實施態樣中，處理 器15可從第一階段之該等參考電力資訊93中，同時選取多個參 考電力資訊，並設定其為第二階段之多個待測電力資訊。 處理器15接著繼續對電器31之電力訊號2取樣，以獲得一新 增電力資訊944’並將新增電力資訊944與第一階段未被選取之參 考電力資訊932、參考電力資訊933及參考電力資訊934，設定為 第二階段之複數個參考電力資訊94。換言之，第二階段之該等參 φ 考電力資訊包含參考電力資訊932、參考電力資訊933、參考電力 資訊934及新增電力資訊944。須說明者，於此實施例中，由於處 理器15僅選取-筆第一階段之參考電力資訊作為第二階段之待測 電力資訊，因此處理器15須再對電器31之電力訊號2取樣一次 以作為第二階段之新增電力資訊。倘若處理器15選取了多筆第一 階段之參考電力資訊作為第二階段之待測電力資訊，處理器15便 須再對電器31之電力訊號2取樣同樣的數目，以作為多筆新增電 力資訊。 接著，處理器15計算第二階段之該等參考電力資訊94之一統 計特性。類似的，處理器15判斷第二階段之待測電力資訊（亦即 參考電力資訊931)是否落入該等參考電力資訊94之統計特性所 界定之一機率分佈範圍内，此機率分佈範圍可為該統計特性之平 均值冑異數或疋其他統計參數進行界定，但i不揭限於此處所 揭露。 最後，處理器15根據第一階段之比較結果及第二階段之比較結 果，判斷第二階段之該等參考電力資訊94是否屬於一穩定狀態。 201224473 由於穩定與不穩定的電力特徵具有不同的統計特性，因此，當第 一階段與第二階段之比較結果均呈現待測電力資訊落入參考電力 資訊之統計特性所界定冬機率分佈範圍内，則可以合理推斷電器 31已處於穩定狀態。此時，處理器15便可將第二階段之該等參考 電力資訊94設定為電器31之電力特徵。此時，第一實施例之電 力特徵辨識裝置1對於該電器3之訓練業已完成。 需說明者，於其他實施態樣中，本領域具通常知識者應可輕易 推斷本發明之電力特徵辨識裝置1可根據更多個階段之比較結果 作為判斷該電器3是否處於穩定狀態，並非僅侷限於此實施例所 揭露之二個階段。 透過第一實施例之配置及運作，本發明之電力特徵辨識裝置1 可根據比對待測電力資料與參考電力資料之統計特性，判斷一電 器是否已處於穩定狀態。據此，將能有效解決習知技術中，依照 使用者經驗判斷各電器之電力訊號是否穩定，造成使用者操作上 的不便利或是電器學習上的不明確等問題。 本發明之第二實施例亦為一電力特徵辨識裝置1，其說明亦請參 考第1圖及第3圖。第3圖係描繪第二實施例之一電力訊號取樣 示意圖。本實施例主要用以闡述電力特徵辨識裝置1如何於電器 辨識階段辨識出電力訊號4之一即時電力特徵。當對電器組3之 所有電器分別進行訓練完成後，電力特徵辨識裝置1於後續便能 對電力迴路9上之電器組3進行監視。於監視階段，電力特徵辨 識裝置1持續地接收電力迴路9上之電力訊號4,並持續地辨識電 力訊號4之電力特徵，以提供使用者電器組3之即時電力資訊。 201224473 處理器15於收到來自電力迴路9之電力訊號4之後，設定一第 一階段之一取樣時間間隔T及一預設取樣次數β取樣時間間隔係 用以決定每隔多少時間間隔對電力訊號4進行—次取樣，而預設 取樣次數係用以對電力訊號4每間隔該取樣時間間隔連續取樣幾 次。在並非用以限定本發明之前提下，為更明確說明本實施例， 以下將假設第一階段之預設取樣次數為4次進行說明。 首先，處理器15每隔取樣時間間隔Τ對電力訊號4取樣，以各 自得到第一階段之一參考電力資訊，直至該等參考電力資訊之一 數目等於預設取樣次數（即4次）為止。換言之，此時每取樣一 次，便得-參考電力資訊，直到得到4個參考電力資訊為止。處 理器15於第-階段獲得複數個（即4個）參考電力資訊％，並將 該等參考電力資訊95儲存於儲存器13。其中，該等參考電力資訊 95係包含-參考電力資訊95]一參考電力資訊松、—參考電力 資訊953及-參考電力資訊州。處理器15將該等參考電力資訊 9—5館存於儲存器13後，再對電力訊心取樣以獲得—第一階段之 一待測電力資訊91a，並將待測電力資訊…儲存於儲存器η。 接著，處理器15計算該等參考電力資訊95之_ 將待測電力資訊91a與該統計特性進行比較 ” :::果。舉例而言，處理…判斷待測電力資訊9= 入该統計特性所界定 疋否洛 一即時電力特徵雷Μ ，以辨識電力訊號4之 判定4右雜 辨識裝置1便可利用該即時電力特徵 疋有新電器啟動、關閉或不正常等情況發生。 率分佈範圍可以栌摅过祕丄 ，、中，该機 料特性之平均值、變異數或是其他統計 201224473 參數進行界定，但並不侷限於此處所揭露。 由於動態地調整該預設取樣次數，將有效地減少抓錯電力特徵 的機會。處理器15可進一步地根據待測電力資訊91a落於該統計 特性所界定之該機率分佈範圍的不同，設定一第二階段之—預設 取樣次數。換言之，處理器15會根據第一階段之比較結果，調整 預設取樣次數。視不同情況而定，第二階段之預設取樣次數可比 第一階段之預設取樣次數大、相同或小。 第二階段會有複數個參考電力資訊96，其數目等於第二階段之 預設取樣次數。處理器15會再次對電力訊號4取樣以得第二階段 之一待測電力資訊92a。接著，處理器15計算第二階段之該等參 考電力資訊96之一統計特性，並將待測電力資訊92a與第二階段 之統計特性進行比較，以再次辨識電力訊號4之另一即時電力特 徵。處理器15藉由週而復始的監視著電力迴路9之電力訊號4， 俾電力特徵辨識裝置1可隨時掌握電力迴路9之電力訊號4之一 即時電力特徵，並判斷是否需要傳送或計算該電力特徵，以有效 地將該即時電力特徵反應至使用者。 為了更明確地闡述如何根據該待測電力資訊落於該統計特性所 數。下文將以一结·—戀思去 ，設定該第二階段之該預設取樣次Qiaomin's good pre-processing technology is easy to waste too much computation or reduce efficiency by transmitting unused packages. There is a clock here to solve the problem of knowing the non-invasiveness of the ground. How to establish a power feature identification device and a method thereof, and the problem of unclear training time and effective power characteristics of the effective loop identification technology is urgently needed for the industry in the field. The purpose of non-* is to provide a power feature identification device to effectively solve the problem caused by the unclear training time of the loop identification technology of the gas and the inability to effectively identify the characteristics of the instant Thunder. The present invention provides a power feature recognition device for continuously receiving a device, a memory, and a processor. The receiver uses the receiver and uses the power signal. The processor is electrically connected to the storage device and the processor is further determined to be - the sampling interval and the number of sampling times. And will sample the power signal to obtain one of the power sources to be tested at this stage: 'The temple power information is stored in the storage. The processor is also configured to sample the power signal every sampling time interval of 201224473, and each of the phases is referenced to the power information until the number of the reference power information is equal to the preset sampling number, and the same The reference power information is stored in the storage. Finally, the processor is configured to calculate a statistical characteristic of the reference power information, and compare the power information to be tested with the statistical characteristic to obtain a comparison result of the phase. To achieve the above object, the present invention also provides a power feature identification method for a device, the device comprising a receiver, a storage and a processor, the method comprising the steps of: (a) causing the receiver to continue Receiving a power signal; (b) causing the processor to set a sampling interval of one phase and a predetermined number of sampling times; (c) causing the processor to sample the power signal to obtain one of the phases to be tested Power information, and storing the power information to be tested in the storage; (d) causing the processor to sample the power signal every sampling interval to obtain power information for each of the phases, until such The number of reference power information is equal to the preset number of samples, and the reference power information is stored in the storage; (e) the processor is configured to calculate a statistical characteristic of the reference power information; (f) The processor compares the power information to be tested with the statistical characteristic to obtain a comparison result of one of the phases. The above objects, technical features, and advantages will be more apparent from the following description. [Embodiment] Hereinafter, a power feature recognizing device and a method thereof according to the present invention will be explained through a plurality of embodiments. It should be noted that since the present invention mainly relates to an apparatus for identifying a power feature and a method thereof, components and steps not directly related to the present invention are omitted in the embodiments and drawings described below. Not shown. In addition, in order to clearly disclose the technical features of the present invention 201224473, the present drawings are used to say that (4) Wei Dingben (4) is exaggerated. This benefit is subject to change. The scope of the request is 'to apply for the first and second diagrams of the first invention of the invention. The first w-line device], the description of which should be combined with the indication of the power circuit 9 = drawing power special identification device 1 applied - sampling schematic. As depicted in the figure... 11, a storage " and a processor 15. The feature recognition device 1 includes a receiver and a receiver 51. The processor 15 is electrically connected to the memory I3 ° . The device 11 is electrically connected to the power circuit 9 and is used to continuously collect power (4) on the power circuit 9? The port 1 is used to continuously connect ^ 9 φ , , ° , and the power signal 2 is from the electrical group 3 electrically connected to the power circuit 9 . = Example of the main _ force characteristics of the Wei set of how to determine the state of electricity _ ... steady state, and then identify - electrical - power characteristics. In the electric (four) training phase, the electric power feature identification device 1 will train the electric appliance, the electric appliance and the electric appliance 35 respectively to learn the electric characteristics of the individual electric appliances of the electric appliance group 3. Taking the training appliance 31 as an example, when the appliance 31 is turned on, the receiver η continuously receives the power signal 2 of the appliance from the power circuit 9. The processor 15 receives the power of the appliance 31 for 2 money's setting I-stage-sampling interval U-preset sampling times. The sampling interval T is used to determine how often the power signal 2 is subjected to the "second time" interval, and the preset sampling money center continuously samples the power signal 2 for several consecutive intervals between the sampling times m. Before the present invention is described, it is assumed that the actual number of the first stage is assumed to be four times. 201224473 As shown in Fig. 2, first, the processor obtains the power signal 2 of the first stage of the power source to be tested, and the power signal 91 to be tested is stored in the memory 13. It should be noted that the power to be tested, | 冤力资 sK 91 in other embodiments of the present invention' may also be the sampling number of the signal 2 of the plurality of power information taken every sampling interval T The device 15 stores the power information to be tested 91 in the memory 13 and samples the power signal 2 of the electrical device 31 at intervals between the sampling times to obtain the first phase of the power-saving battery. Refer to the power information to preset the number of samples. In other words, at this time, every time the sample is taken - the power information is obtained - until four reference powers are received. The processor 5 obtains a plurality of (i.e., 4) moxibustions, and refers to the power information 93 in one stage, and stores the reference powers ^93 in the storage unit 13. The reference power information 93 includes a reference power information 931, a reference +f, and a reference power information 934. Beixinxin 1 reference power information 933 will = electric = 15 calculate the statistical characteristics of the reference power information 93, and the measured power W1 and the reference power information 93 statistical characteristics are compared, and then a comparative fruit Hinder # U example and tongue, processor! 5 can be compared to the probability distribution of the power to be measured into one of the statistical characteristics defined by the statistical characteristics, where = cloth: according to the statistical characteristics of the average, the number of variations or other systems: set '疋' but not brown limited Revealed here. The comparison result obtained by the first-order production ratio Ιλ result obtained here is the total of (4) the basis for whether or not the contactor 31 has completed the training. After σ is too much to obtain the above comparison result, the processor 5 selects the reference power information state from the reference power information 93 of the first stage, and sets it as a second stage of 201224473 a power information to be tested. In other embodiments of the present invention, the processor 15 may select a plurality of reference power information from the reference power information 93 of the first stage, and set the plurality of reference powers to be tested. Power information. The processor 15 then continues to sample the power signal 2 of the appliance 31 to obtain a new power information 944' and adds the power information 944 with the reference power information 932, the reference power information 933, and the reference power that are not selected in the first stage. The information 934 is set to a plurality of reference power information 94 of the second stage. In other words, the second stage of the reference power information includes reference power information 932, reference power information 933, reference power information 934, and new power information 944. It should be noted that, in this embodiment, since the processor 15 only selects the reference power information of the first stage of the pen as the power information to be tested in the second stage, the processor 15 needs to sample the power signal 2 of the electric appliance 31 again. Take the new power information as the second stage. If the processor 15 selects a plurality of first-stage reference power information as the second-stage power information to be tested, the processor 15 needs to sample the same number of power signals 2 of the electrical device 31 as multiple new powers. News. Next, the processor 15 calculates a statistical characteristic of the reference power information 94 of the second phase. Similarly, the processor 15 determines whether the power information to be tested in the second stage (ie, the reference power information 931) falls within a probability distribution defined by the statistical characteristics of the reference power information 94. The probability distribution range may be The average of the statistical properties is different from the other statistical parameters, but it is not disclosed herein. Finally, the processor 15 determines whether the reference power information 94 of the second stage belongs to a steady state based on the comparison result of the first stage and the comparison result of the second stage. 201224473 Because the stable and unstable power characteristics have different statistical characteristics, when the comparison between the first stage and the second stage shows that the power information to be tested falls within the winter machine rate distribution defined by the statistical characteristics of the reference power information, It can be reasonably concluded that the appliance 31 is in a stable state. At this time, the processor 15 can set the reference power information 94 of the second stage to the power characteristics of the appliance 31. At this time, the training of the electric characteristic recognition apparatus 1 of the first embodiment for the electric appliance 3 has been completed. It should be noted that in other embodiments, those skilled in the art should readily infer that the power feature recognition apparatus 1 of the present invention can judge whether the appliance 3 is in a stable state according to the comparison result of more stages, not only It is limited to the two stages disclosed in this embodiment. Through the configuration and operation of the first embodiment, the power characteristic identifying apparatus 1 of the present invention can determine whether an electric motor is in a stable state based on the statistical characteristics of the power data to be measured and the reference power data. Accordingly, it is possible to effectively solve the conventional techniques, and judge whether the electric power signals of the electric appliances are stable according to the user experience, resulting in inconvenience in the operation of the user or ambiguity in the learning of the electric appliance. The second embodiment of the present invention is also a power feature recognition device 1. For the description, please refer to Figs. 1 and 3. Fig. 3 is a diagram showing a power signal sampling of a second embodiment. This embodiment is mainly used to explain how the power feature identification device 1 recognizes an instant power feature of the power signal 4 during the appliance identification phase. After the training of all the appliances of the electrical group 3 is completed, the power signature identifying device 1 can subsequently monitor the electrical group 3 on the power circuit 9. During the monitoring phase, the power feature identification device 1 continuously receives the power signal 4 on the power circuit 9 and continuously recognizes the power characteristics of the power signal 4 to provide instant power information of the user's electrical group 3. 201224473 After receiving the power signal 4 from the power circuit 9, the processor 15 sets a sampling interval T and a preset sampling time β sampling interval for determining the power signal every other time interval. 4 - sampling is performed, and the preset number of samplings is used to continuously sample the power signal 4 at intervals of the sampling interval. Before the present invention is not limited, in order to more clearly explain the present embodiment, the following description will be made assuming that the preset number of sampling times in the first stage is four. First, the processor 15 samples the power signal 4 every sampling interval to obtain the reference power information of one of the first phases until the number of the reference power information is equal to the preset number of samples (i.e., 4 times). In other words, every time you sample this time, you will get the reference power information until you get 4 reference power information. The processor 15 obtains a plurality (i.e., four) of reference power information % at the first stage, and stores the reference power information 95 in the memory 13. The reference power information 95 includes - reference power information 95] a reference power information loose, - reference power information 953 and - reference power information state. The processor 15 stores the reference power information 9-5 in the storage device 13, and then samples the power signal core to obtain one of the first stage power information to be tested 91a, and stores the power information to be tested in the storage. η. Next, the processor 15 calculates the reference power information 95 to compare the power information to be tested 91a with the statistical characteristic. ::: Fruit. For example, the process determines the power information to be tested 9 = enters the statistical characteristic Defining the 疋 洛 洛 an instant power characteristic thunder to identify the power signal 4, the right miscellaneous identification device 1 can use the instant power feature to start, close or abnormally new appliances. The rate distribution range can be丄 丄 丄 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , The opportunity of the power feature is captured. The processor 15 may further set a second phase of the preset sampling times according to the difference of the probability distribution range defined by the power information 91a to be measured by the statistical characteristic. In other words, the processor 15 will adjust the preset sampling times according to the comparison result of the first stage. Depending on the situation, the preset number of sampling in the second stage can be compared with the first stage. The preset sampling times are large, the same or small. In the second stage, there will be a plurality of reference power information 96, the number of which is equal to the preset sampling number of the second stage. The processor 15 will sample the power signal 4 again to obtain the second stage. A power information to be tested 92a. Next, the processor 15 calculates a statistical characteristic of the reference power information 96 of the second stage, and compares the power information to be tested 92a with the statistical characteristics of the second stage to identify the power signal again. Another instant power feature of the processor 4. The processor 15 monitors the power signal 4 of the power circuit 9 by repeating the cycle, and the power feature identification device 1 can grasp the instantaneous power feature of the power signal 4 of the power circuit 9 at any time, and determine whether or not The power feature needs to be transmitted or calculated to effectively reflect the instantaneous power feature to the user. In order to more clearly explain how the power information according to the test is based on the statistical characteristics, the following will be based on a knot. Go, set the preset sampling time of the second stage
界定之該機率分佈範圍的不同Defining the difference in the probability distribution range
小於該第二變異數， 若待測電力資訊91 a落於該等參考電力資訊9 5 率範圍，且此機率範圍為大於該第一變異數且 ，表示待測電力資訊9la為已知特性，於是處 201224473 理器15將τ-階狀㈣設㈣錄設定與切段线預設取樣 次數相同（亦即，第二階段之預設取樣次數等於第—階段之預設 取樣次數，為4次）。 接著，處理器15將設定參考電力資訊952、953、gw及待測電 力資訊91a設為第二階段之複數個參考電力資訊％，該等參考電 力資訊96之數目等於第二階段之預設取樣次數。於下一時間點（經 過-個取樣時間間隔），處理器15再對電力訊號4取樣以獲得 # —待測電力資訊92a。處理器15接著計算第二階段之該等參^電 力資訊96之-統計特性，並將之與待測電力資訊似比較以得第 二階段之-比較結果。假設第二階段之比較結果為待測電力資訊 92:落於-機率分布範圍（例如大於該第二變異數），表示待測電 力資訊92a可能是同一個電器的特徵。於是’處理器15將下—階 段（亦即第三階段）之倾取樣次數設定為大於本階段（亦即第 二階段）之預設取樣次數。If the power information to be tested 91 a falls within the range of the reference power information, and the probability range is greater than the first variation, it indicates that the power information to be tested 9la is a known characteristic. Then, at 201224, the processor 15 sets the τ-step (four) setting (four) recording setting to the same number of preset sampling times as the segment line (that is, the preset sampling number of the second stage is equal to the preset sampling number of the first stage, which is 4 times. ). Then, the processor 15 sets the set reference power information 952, 953, gw and the power information to be tested 91a as a plurality of reference power information % of the second stage, and the number of the reference power information 96 is equal to the preset sampling of the second stage. frequency. At the next time point (via a sampling interval), the processor 15 samples the power signal 4 to obtain # - the power information to be tested 92a. The processor 15 then calculates the statistical characteristics of the reference power information 96 of the second stage and compares it with the power information to be tested to obtain the second stage-comparison result. It is assumed that the comparison result of the second stage is the power information to be tested 92: falling within the range of probability distribution (for example, greater than the second variation), indicating that the power information to be tested 92a may be a feature of the same electrical appliance. Thus, the processor 15 sets the number of down samplings of the lower stage (i.e., the third stage) to be larger than the preset number of samples of the current stage (i.e., the second stage).
欠接著’處理器15設定參考電力資訊心扮州及待測電力 資訊9la、92a設為第三階段之複數個參考電力資訊97，該等參考 電力資訊97之數目等於第三階段之預設取樣次數（例如5次）。 :再下時間點，處理器j 5再對電力訊號4取樣，以獲得—待測 電力資訊93a。處理器15接著計算第三階段之該等參考電力資訊 97之-統計特性，並將之與待測電力資訊—比較以得第三階段 之-比較結果。若第三階段之比較結果為待測電力資訊％落於— 機率分布範圍（例如小於該第一變異數），表示待測電力資訊％ 可能是新進電器的特徵，於是處理器15將下—階段（亦即第四階 13 201224473 段）之預設取樣次數設定為小於本階段（亦即第三階段）之預設 取樣次數。於下一階段（亦即第四階段），處理器15便會設定待 測電力資訊91a、92a、93a第四階段之複數個參考電力資訊％， 該等參考電力資訊98之數目等於第四階段之預設取樣次數（例如 3次）〇 需說明者係，上述之說明僅用以更明確地表示本發明之實施態 樣’並非用以舰本發明。本領域具通常知識應可㈣置換其他 參數界㈣統計特性之該機率分佈範圍，並㈣設定下—階段之 該預設取樣次數，而非僅如第3圖所示之態樣。 透過第二實施例之配置及運作，本發明之電力特徵辨識裝置ι 可根據比對待測電力資料與參考電力資料之統計特性，隨時掌握 電力迴路9之電力訊號4之—即時電力特徵。據此，將能有效解 決習知技射，浪費過多的計算量或相傳送過多無㈣封包而 降低效率等問題。 本發明之第三實施例為一種用於-裝置之電力特徵辨識方法’ 铺置包含-接收器一儲存器及_處理器，其中，該接 ==處理器可分別為第—實施例之接收器"、儲存器U 換言之，該裝置可以為第—實施例中之電力特徵辨 此外，第三實施例所描述之電力特徵辨識方法可由— 產品執行，當《置載人該電腦程式產品，並 ζ 品所包含之複數個指令後，即可 ⑽&式產 辨m 第二實施例所述之電力特徵 辨識方法。前述之電腦程式產品 ㈣ 仔於電腦可蟥取記錄媒體 201224473 】如唯項δ己憶體(read only mem〇ry ; R〇M)、快閃記憶體、軟 、更碟光碟、隨身碟、磁帶、可由網路存取之資料庫或熟習 項技农者所習知且具有相同功能之任何其它儲存媒體中。 第4Α-4Β⑸系描綠第三實施例之流程圖。首先，由步驟S301 y “接收ϋ持續地接收―電力訊號。接著於步驟㈣2，令該處理 器6又疋-第_階段之一取樣時間間隔及一預設取樣次數。步驟 咖係用以令該處理器對該電力訊號取樣以得-第-階段之一待 #測電力資訊，再由步驟讓令該處理器將該待測電力資訊儲存於 ο:存ϋ於步驟請3之候，由步驟S3〇5令該處理器每隔該取 樣^間隔對該電力訊號取樣，以各自得該第-階段之-參考電 力貝讯’直至該等參考電力f訊之—數目等於該預設取樣次數， 再由步驟S306令該處理器將該等參考電力資訊儲存於該健存器。 接著由步驟S307令該處理器計算該等參考電力資訊之—統計特 性’並由步驟請8令該處理器判斷該待測電力資訊落入該統計特 性所界定之-範_，以得該第—階段之—比較結果。 =步驟S綱之後，步驟請9令該處理器選取前一階段之該等 =電力資訊之至少-個。於步驟請9之後，由步驟_令該 處理益設定該被選取之至少一參考電力資訊為本階段之一待測電 力資訊，以及由步驟S31I令該處理器對該電力訊號取樣以得至少 一新增電力資訊。於步驟削及灿之後，由步驟S312令該處 理盗將該至少—新增電力資訊與前_階段未被選取之魅少一參 考電力資訊，設定為本階段之複數個參考電力資訊。於步驟仙 之後，由步驟則令該處理器計算本階段之該等參考電力資戒之 15 201224473 一統計特性。於步驟S313之後，由步驟S314令該處理器判斷本 階段之該待測電力資訊落入本階段之該統計特性所界定之一範圍 内0 於步驟S3 14之後，由步驟S3 15令該處理器判斷本階段之該等 參考電力資訊屬於-穩定狀態。其中，步驟則係根據本階段所 獲得之比較結果與先前階段所獲得之比較結果，判斷該等參考電 力資訊屬於-穩定狀態。若判斷結果為是，則結束電器訓練。若 判斷結果為否，則返回至步驟S3〇9遞迴處理。 除了上述步驟，第三實施例亦能執行第一實施例所描述之所有 操作及功能，所屬技術領域具有通常知識者可直接瞭解第三實施 例如何基於上述m例以執行此#操作及魏，故不賛述。 透過第三實施例之流程說明，本發明之電力特徵辨識方法可根 據比對待測電力資料與參考電力資料之料雜，觸—電器a 否已處於穩定狀態。據此，將能有效解決習知技術中依昭使= 者經驗判斷各電n之電力訊號是否穩定，造成使㈣操作上的不 便利或是電器學習上的不明確等問題。 、本發明之第.四實施例亦為—種用於—裝置之電力特徵辨識方 法’該裝置包含-接收器、一儲存器及一處理器其令，該接收 15、該儲存器及該處理器可分別為第—實施例之接收器U、錯存 器13及處理H 15。換言之，該裝置可以為第—實施财 徵辨識裝置I。 # 此外，第四實施例所描述之電力特徵辨識方法亦可由_電腦程 201224473 式產品執行，當該裝置載人該電腦程式產品，並執行該電腦程式 產品所包含之複數個指令後，即可完成第三實施例所述之電力特 徵辨識方法。前述之電肺式產品可儲存於電腦可讀取記錄媒體 中例如唯項記憶體(read only memory ; R〇M)、快閃記憶體、軟 碟、硬碟、光碟、隨身碟、磁帶、可由網路存取之資料庫或熟習 此項技藝者所習知且具有相同功能之任何其它儲存媒體中。 第5圖係描繪第四實施例之流程圖。首先，由步驟S4〇1令該接 收器，持續地接收一電力訊號。於步驟S4〇2中，令該處理器設定 一取樣時間間隔及一預設取樣次數。於步驟34〇2後，由步驟 令該處理器每隔該取樣時間間隔對該電力訊號取樣，以各自得一 參考電力資訊，直至該等參考電力資訊之一數目等於該預設取樣 次數，並由步驟84〇4令該處理器將該等參考電力資訊儲存於該儲 存器。於步驟S403後，由步驟S405令該處理器對該電力訊號取 樣以得一待測電力資訊，並由步驟S406令該處理器將該待測電力 二貝Λ儲存於該儲存器。於步驟S4〇7中，令該處理器計算該等表考 電力資訊之一統計特性。之後，由步驟S4〇8令該處理器判斷該待 測電力資訊落入該統計特性所界定之一範圍内，以得該第—階段 之一比較結果。於步驟S409中，令該處理器根據該比較結果，改 變忒預設取樣次數，並返回至步驟S4〇3遞迴處理。 除了上述步驟，第四實施例亦能執行第二實施例所描述之所有 才呆作及功能，所屬技術領域具有通常知識者可直接瞭解第四實扩 例如何基於上述第二實施例以執行此等操作及功能，故不贅述施 透過第四實施例之配置及運作，本發明之電力特徵辨識方法可 17 201224473 根據比對待測電力眘祖你 雷考電力㈣之料雜，_掌握’ 電力訊號之一即時雷.姓 旱握該 浪費過多料算量< β㈣，聽有效解決f知技術中， 題。 &疋因傳送過多無用的封包而降低效率等問 =特別加以說明者係，上述之實施㈣料例舉本 二以及_本發明之技術特徵，並非用來限制本二 巳可任何熟悉此技術者可輕易完叙改變或均等性之安排均屬 於本發月所主張之範@ ,本發明之權㈣護㈣應 圍為準。 月f j祀 【圖式簡單說明】 第1圖係描繪電力特徵辨識裝置1應用於一電力迴路9之示意 圖； 第2圖係描繪第一實施例之一電力訊號取樣示意圖 第3圖係描繪第二實施例之一電力訊號取樣示意圖 第4Α·4Β圖係描繪第三實施例之流程圖；以及 第5圖係描繪第四實施例之流程圖。 【主要元件符號說明】 11 :接收器 15 :處理器 3 :電器組 9 :電力迴路 1 :電力特徵辨識裝置 13 :儲存器 2、4:電力訊號 31 ' 33、35 :電器 91 ' 91a、92a、93a、94a :待測電力資訊 201224473After the processor 15 sets the reference power information, the state and the power information to be tested 9la, 92a are set to the third stage of the plurality of reference power information 97, and the number of the reference power information 97 is equal to the preset sampling of the third stage. The number of times (for example, 5 times). At the next time point, the processor j 5 samples the power signal 4 to obtain the power information 93a to be tested. The processor 15 then calculates the statistical characteristics of the reference power information 97 of the third stage and compares it with the power information to be tested to obtain the third stage-comparison result. If the comparison result of the third stage is that the power information to be tested falls within the probability distribution range (for example, less than the first variation), it indicates that the power information to be tested may be a feature of the new incoming device, and then the processor 15 will be in the next stage. The preset sampling number of the fourth stage 13 201224473 is set to be less than the preset sampling number of this stage (ie, the third stage). In the next stage (ie, the fourth stage), the processor 15 sets a plurality of reference power information % of the fourth stage of the power information 91a, 92a, 93a to be tested, and the number of the reference power information 98 is equal to the fourth stage. The preset number of samplings (e.g., three times) is not required to be described, and the above description is only used to more clearly show that the embodiment of the present invention is not used in the present invention. The general knowledge in the field should be (4) to replace the probability distribution range of the other parameter boundaries (4) statistical characteristics, and (4) to set the preset sampling times of the next-stage, rather than just the aspect shown in Figure 3. Through the configuration and operation of the second embodiment, the power feature identification device ι of the present invention can grasp the instantaneous power feature of the power signal 4 of the power circuit 9 at any time based on the statistical characteristics of the power data to be measured and the reference power data. According to this, it is possible to effectively solve the conventional technique, waste too much calculation amount or transmit too many (4) packets to reduce efficiency and the like. A third embodiment of the present invention is a power feature identification method for a device. A storage-receiver-reservoir and a processor are disposed, wherein the connection== processor can be respectively received by the first embodiment. In other words, the device can be identified as the power feature in the first embodiment. In addition, the power feature identification method described in the third embodiment can be executed by the product, when the "personality of the computer program product, After the plurality of instructions included in the product, the power feature identification method described in the second embodiment can be produced by the (10) & The aforementioned computer program products (4) can be used to capture recording media 201224473] such as only δ 忆 recall (read only mem〇ry; R〇M), flash memory, soft, disc, flash drive, tape A database accessible by the network or any other storage medium known to the skilled worker and having the same function. 4th-4Α(5) is a flow chart of the third embodiment of the green. First, the power signal is continuously received by the receiving device in step S301 y. Next, in step (4) 2, the processor 6 is further configured to sample the time interval and a preset sampling frequency. The processor samples the power signal to obtain the power information of the first stage, and then causes the processor to store the power information to be tested in the step ο: Step S3〇5 causes the processor to sample the power signal every interval of the sampling interval, so that the number of the first-phase-reference power signals is up to the number of the predetermined sampling times. And then, in step S306, the processor stores the reference power information in the payload. Then, in step S307, the processor calculates the statistical property of the reference power information and the processor is configured by the step 8 Determining that the power information to be tested falls within the __ defined by the statistical characteristic, so as to obtain the comparison result of the first stage. After the step S, the step is to order the processor to select the previous stage. At least one of the power information. In the steps After 9th, the process determines that the selected at least one reference power information is one of the power information to be tested in the current stage, and the processor samples the power signal to obtain at least one new power by step S31I. After the step is cut, the step S312 causes the processing thief to set the at least one of the new power information and the unselected enchantment reference power information of the previous stage to set a plurality of reference power information for the stage. After the step, the processor causes the processor to calculate the statistical characteristics of the reference power resources of the current stage 15 201224473. After the step S313, the processor determines, by the step S314, the power information to be tested at the current stage. Within the range defined by the statistical characteristic of the current stage, after step S3 14 , the processor determines, by step S3 15 , that the reference power information of the current stage belongs to a stable state, wherein the step is based on the present The comparison result obtained in the stage and the comparison result obtained in the previous stage determine that the reference power information belongs to a stable state. If the judgment result is yes, then the knot Electrical training. If the result of the determination is no, the process returns to step S3 to return to the processing. In addition to the above steps, the third embodiment can also perform all the operations and functions described in the first embodiment, and those skilled in the art have ordinary knowledge. It can be directly understood how the third embodiment performs the operation and the Wei based on the above m example, so it is not praised. Through the flow description of the third embodiment, the power feature identification method of the present invention can calculate the power data and reference according to the ratio. The power data is mixed, and the touch-electrical appliance a is in a stable state. According to this, it will be able to effectively solve the problem of whether the power signal of each electric power is stable in the conventional technology, and the operation of the electric power signal is stable. Convenience or ignorance in electrical learning. The fourth embodiment of the present invention is also a power feature identification method for a device. The device includes a receiver, a memory, and a processor. The receiver 15, the memory and the processor are respectively the receiver U, the register 13 and the processing H 15 of the first embodiment. In other words, the device can be the first implementation of the identification device I. The power feature identification method described in the fourth embodiment can also be executed by the computer program 201224473. When the device carries the computer program product and executes a plurality of instructions included in the computer program product, The power feature identification method described in the third embodiment is completed. The above-mentioned electric lung type product can be stored in a computer readable recording medium such as a read only memory (R〇M), a flash memory, a floppy disk, a hard disk, a compact disk, a flash drive, a magnetic tape, or the like. A database of network access or any other storage medium known to those skilled in the art and having the same functionality. Figure 5 is a flow chart depicting a fourth embodiment. First, the receiver is continuously received by a step S4〇1 to receive a power signal. In step S4〇2, the processor is configured to set a sampling time interval and a preset sampling number. After step 34〇2, the processor causes the processor to sample the power signal every sampling interval to obtain a reference power information until the number of the reference power information is equal to the preset sampling number, and The processor stores the reference power information in the storage by step 84〇4. After step S403, the processor determines the power signal to obtain a power information to be tested in step S405, and causes the processor to store the power to be tested in the memory in step S406. In step S4, the processor is caused to calculate a statistical characteristic of the power information. Then, the processor determines, by the step S4〇8, that the power information to be tested falls within a range defined by the statistical characteristic, so as to obtain a comparison result of the first phase. In step S409, the processor is caused to change the preset number of sampling times according to the comparison result, and returns to step S4〇3 to return the processing. In addition to the above steps, the fourth embodiment can also perform all the functions and functions described in the second embodiment, and those skilled in the art can directly understand how the fourth embodiment is based on the above second embodiment to perform this. The operation and function are not described, and the configuration and operation of the fourth embodiment are not described. The power feature identification method of the present invention can be used according to the ratio of the electric power to be tested. One of the instant mines. The surname is a dry grip that wastes too much material calculation < β (four), and listens effectively to the problem of knowing the technology. & 降低 传送 传送 传送 传送 传送 传送 传送 传送 传送 传送 传送 传送 传送 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = The arrangement that can easily complete the change or the equality is the Fan @ claimed by this month, and the right of the invention (4) (4) should be subject to the standard. Ffj祀 [schematic description of the drawings] Fig. 1 is a schematic diagram showing the application of the power characteristic identification device 1 to a power circuit 9; Fig. 2 is a diagram showing the power signal sampling of the first embodiment. One embodiment of the power signal sampling diagram is a flowchart of the third embodiment; and FIG. 5 depicts a flow chart of the fourth embodiment. [Description of main component symbols] 11 : Receiver 15 : Processor 3 : Electrical group 9 : Power circuit 1 : Power characteristic identification device 13 : Memory 2 , 4 : Power signal 31 ' 33 , 35 : Electrical device 91 ' 91a , 92a , 93a, 94a: power to be tested 201224473
93、931、932、933、934、94 :參考電力資訊 944 :新增電力資訊 95、951、952、953、954、96、97、98 :參考電力資訊 T:取樣時間間隔 1993, 931, 932, 933, 934, 94: Reference Power Information 944: New Power Information 95, 951, 952, 953, 954, 96, 97, 98: Reference Power Information T: Sampling Interval 19